MS-QA_06-2019 rev5

ANSI/AARST MS-QA 2019 11 Radon Measurement Systems Quality Assurance Requirements For All Methods Informative advisory —Complete measurement system calibration should be performed before initial use or following maintenance, repair or any other changes that could affect performance (ASTM D7282 2006), and as required by method-specific sections of this standard. 3.5 Default Minimum Criteria for Warning and Control Limits—Duplicates and Spikes In all cases, warning and control limits shall be equal to or more stringent than the minimums provided in this standard. When applying more stringent minimums, default warning and control limits shall be updated using the results of QC checks following statistically defensible control limit generation algorithms, such as described in the MS-QA Companion Guidance Section B . 3.5.1 Duplicate and Comparison Check Target for Imprecision Note—The default interim limits presented in this standard were derived assuming a conservative “ in- control ” target for imprecision that is demonstrated by the average relative percent difference (RPD) of duplicate and comparison checks being not greater than: a) 14% in an environment with a radon concentration ≥ 4 pCi/L (150 Bq/m 3 ), and b) 25% in an environment with a radon concentration between 2 and 4 pCi/L (75-150 Bq/m 3 ). 3.5.2 Warning and Control Limits for Duplicate and Comparison Checks Minimum warning and control limits required in this standard are: a) In an environment with a radon concentration ≥ 4 pCi/L (150 Bq/m 3 ) — The warning limit is reached when a duplicate pair or comparison check exhibits an RPD ≥ 28%. — The control limit is reached when a duplicate pair or comparison check exhibits an RPD ≥ 36%; b) In an environment with a radon concentration between 2 and 4 pCi/L (75-150 Bq/m 3 ) — The warning limit is reached when a duplicate pair or comparison check exhibits an RPD ≥ 50%. — The control limit is reached when a duplicate pair or comparison check exhibits an RPD ≥ 67%; and c) If the average of a duplicate pair or comparison check is less than 2 pCi/L(75 Bq/m 3 ), the warning limit is reached when there is a difference between the two results of more than 1.0 pCi/L (37 Bq/m 3 ), or if both results are not less than the minimum detectable concentration (MDC). 3.5.3 Spike Target for Measurement System Accuracy Note—The default interim warning and control limits presented here for spikes were derived assuming in- control operations produce a distribution of spike results centered on zero, with a sample standard deviation of 10%. 3.5.4 Warning and Control Limits for Spikes Any single spike result exhibiting a relative percent error (RPE) (or individual percent error [IPE]) outside the range of 0 ± 20% has exceeded the warning limit, and any single spike result outside the range of 0 ± 30% has exceeded the control limit. 3.5.5 Informative Background for Understanding Warning and Control Limits Limits in this standard are based upon a conservative estimate of industry performance. Most systems will exhibit smaller imprecision and bias errors. It is important to recognize that for limits generated based on actual RPD and RPE performance of a stable and in-control system, the warning limit is expected to be exceeded by chance 5% of the time, and the control limit is expected to be exceeded by chance 1% of the time; therefore, investigation after a failed check may not lead to corrective action . However, if more than 5% of QC checks fall outside warning limits or more than 1% fall outside control limits, an investigation should result in either corrective action or, when applying more stringent limits, regeneration of the limits that reflect actual in-control operations.